epcc workshop 2019 - 15 epcc workshop - may 12 – 15, 2019 - … · 2016. 11. 7. · objectives...
TRANSCRIPT
Open Source Applications
for Control Centers
Challenges and Opportunities
Robin Podmore
EPCC 12th International Workshop
Bedford Springs, PA
June 2-5, 2013
www.incsys.com
Objectives
• Open source product development challenges,
with a specific focus on the smart power grid.
• Engage developers, grid operators, and
companies that provide products and services to
the industry.
• Manage the transition from prototype to market.
• Requirements to be used by the code
developers to ensure successful delivery
"This material is based upon work performed by IncSys
and PowerData and has been supported by the
Department of Energy [National Energy Technology
Laboratory] under Award Number(s) DE-OE0000488.”
DARPA vs. ARPA-E
• “DOD understands that big systems are very important
and very hard to manage.
• Real difference in performance between a weapons
system and its predecessors was usually not the
consequence of one or two or three scientific advances
or technological capabilities.
• But the synergistic effect of one hundred, two hundred or
three hundred advances each of which alone was
relatively insignificant.”
John Alic. Bipartisan Policy Center March 28th, 2012 event on
“Energy Innovation at the Department of Defense: Assessing the Opportunities”:
DARPA vs. ARPA-E
• That is true of complex military systems and it is true of
our energy system which is sprawling and complex and
messy and hard to understand.
• Breakthroughs by themselves do not contribute to
innovation until they work their way through in these
systems.
• If you are guided by feedback from users, from the war
fighters and from the people who do training for the
military, then it helps you figure out what your agenda for
R&D in the future should be.
John Alic. Bipartisan Policy Center March 28th, 2012 event on
“Energy Innovation at the Department of Defense: Assessing the Opportunities”:
Communication Barriers
Utility
System
Operator
Utility
Operations
Engineer
Power
System
Application
Developer
Software
Developer
Substations
Breakers
Switching Orders
EMS, DMS User
Buses
Branches
Power Flow
Transient Stability
PSSE, PSLF
System Protection
FORTRAN
MATLAB
VB
Decoupled Power Flow
Jacobian Matrix
Eigen Vectors
Transient Saliency
PSSE, PSLF
EMS, DMS Apps
Java
C++
OO Programming
OO data base
Threads
Network
Cyber Security
Linux
CIM XML
EMS, DMS SW.
Need for Simulation Test Beds
IEEE 118 Bus (prototype)
• Bus Branch Models
• Single User
• Study Application
• User driven
• Input: user defined files
• 118 Buses
• Runs for select snapshots
• Not mission critical
• Not high availability
• Runs under select conditions
• Stand alone application
Real World (market)
• Breaker Switch Models
• Many Users
• Real-time applications
• Event driven
• Input: real-time measurements
• 50,000 Buses
• Run for days
• Mission critical
• High availability
• Runs under all conditions
• Part of a complex system
PowerSimulator Training Sites
Generic models
Custom models
Station
Customized
from PSSE
Data
• Familiar layouts
• Custom Names
• Familiar Switching
• Transfer Busses
• Suitable for:
•EOP-005
•PER-005
•Blackstart Drills
Note: all equipment names have
been deleted from this display
PG&E Restoration Training • Regional drill by PG&E to train with neighbors
• All attendees work to restore PG&E Custom
• PowerSimulator model from irregular islands
• TOP/TOs, GOP/GOs present
• Takes place annually in Nov or Dec
CETAC Restoration Drill
Northern California
WECC, CAISO, PG&E,
SMUD, TID, MID, WAPA
Southern California
WECC, CAISO, SCE, LADWP,
SDGE, CFE
• 5 Sessions each April/May
• 10th Anniversary 2013 – no more paper drills!
• 500 operators total from all major utilities in CA
Synchroscope
Usability Tests at PNNL EIOC
• Instructor lead NERC CEH Class of 16 Operators.
• Control and Experimental Groups
Operator Decision Model
Massive On-Line Grid Games
(MOGGs)
• Cloud based hosting.
• Web clients.
• Hundreds of on-line users.
• Interconnection wide models.
• Millions of real-time
measurements.
• Simulation updates of 100
times per second.
• Models cascading outages of
interconnections.
• Agents play roles of
neighboring operators.
Reliability Tools for SA
PowerSimulator Architecture PACE – Power Application Computing Environment
by IncSys and PowerData
Common Information Model (with PowerData Real Time Database)
Power Integrator – Application Integration Framework
EPRI
OTS
Data Engineer Power Visuals User Interface
(Web Enabled)
TANGO Contingency
Analysis
PACE Infrastructure Application or System
DNP3/
ICCP
Model
Importers
System
Restoration
Navigator
(EPRI)
Interfaces for Applications
• Import models: CIM/XML, PSSE, PSLF
• Export power system model, API and CSV file
• Export power system state, API and CSV file
• Accept power system controls, API.
• Exchange real-time data: DNP3 and ICCP.
Object Oriented Views of CIM
Encapsulation to Hide Details
Abstraction
Different Views for Different Users
Ref: Grady Booch, “Object Oriented Design”
Software Developers
Must Create the Illusion of Simplicity
Ref: Grady Booch, “Object Oriented Design”
TANGO
Transient Analysis of Grid Operations
LINKNET Parallel Sparse Matrix Solution
Routines
Transient Analysis of Grid Operations
(TANGO)
PowerSimulator
• Combined transient and long
term dynamic simulation
models
• Open source
• Extensible model library
• Tailored for Operator Training
Simulators
• Designed for HPC
• Needed:
“A FEW GOOD
ENGINEERS”
– Java Programming
– VB and Power System
Apps
Developer Requirements
• Prototype applications:
– Test with Cascadia and PALCO models.
– Test with real operators.
– Embed IncSys open source VB code (Option).
• Production applications:
– Test with actual system models.
– Background checks and NDAs are required.
– Test with real operators in the loop.
– Embed IncSys/PowerData Java code (Option).
• Understand how an expert operator currently handles the
problem being studied.
Benefits Simulation Test Beds
• Allow new control methods, analytical applications
and visualization techniques to be thoroughly
tested.
• Feedback from system operators and trainers can
guide the development.
• Vital step to go from prototype to on-line
deployment.
• Time to go from Prototype to Production can be
accelerated from years to months.
https://github.com/powerdata select “Tango”